ADD-QIA: An Adaptive Data Deduplication Framework Based on Quantum Immune Algorithm

Abstract Cloud computing has become the backbone of modern data management, yet the exponential growth of unstructured data from IoT devices, virtual machines, and enterprise systems has created excessive redundancy. Conventional deduplication techniques, such as fixed-size and content-defined chunking, either miss shifted duplicates or impose heavy computational overhead, limiting their scalability. Bio-inspired approaches (GA, PSO, IA) introduce adaptability but suffer from slow convergence and suboptimal trade-offs between deduplication ratio, execution time, and memory usage. To address these gaps, this paper proposes ADD-QIA, an Adaptive Data Deduplication framework based on a Quantum Immune Algorithm. By combining quantum-inspired probabilistic encoding with immune clonal selection, ADD-QIA dynamically adjusts chunking strategies to workload variations. Extensive evaluation on VM snapshots, enterprise backups, and cloud traces demonstrates that ADD-QIA achieves a deduplication ratio of 5.3:1, reduces execution time by 20%, and lowers memory usage by 15%, while sustaining throughput above 300 MB/s. The strength and scalability of the method is statistically validated. These findings make ADD-QIA a viable and scalable model of removing redundancy in the cloud.